On/Off switch by a voltage threshold - Possible SSR use

[fizzy123]

First, let me correct my error. The size of the cap should have been 220µF, or preferably 470µF.

The reasoning is this. At the beginning of the pulse, the cap is discharged. When the pulse arrives the cap will start to charge via the 560ohm resistor. The voltage across it will rise exponentially with time and the final charge would be 6V. But it will only be charging for 10ms and by that time we want the voltage to have increased to no more than 1.5V, so the SSR won't fire. To get to that voltage in 10mS means the time constant (T) needs to be 50ms.

The time constant (T) is given by C x R, where C is in microfarads and R is in megohms. and T is in seconds. So CR= T =50ms. Transposing for C gives C = 50ms/R. I had taken R as 560 but R is actually the input impedance of the voltage source, and that is the 560ohms resistor and the 560ohms input resistance in parallel., which is 280ohms.

So the capacity C will work out at 0.05sec/0.000280megohms = 178.5µF. The minimum practical value is 220µF, but 470µF would be better. C could be any elecrolytic, but preferably could be a low ESR type

I hope that answers your question.

Thanks man. Yes answers my question but few follow up questions.

1. How do you know that time constant T is 50ms for a 10ms delay?
2. I do not see anywhere that how the source voltage of 12v has any influence. What am I missing?
3. I think you said the 560ohm resistor needs to be in series at the SSR input but above you are saying put it in parallel. If you can do a simple circuit diagram that will erase the confusion.
4. And finally if the SSR input resistance is more or less than 560ohms, how it will affect the resistor and capacitor selection?
5. What capacitor voltage rating is safe for my use; I assume at least 16v as source will be 12v?

Sorry for too many questions. No intension to ride your nerves

[pebe]

OK. I'll do you a circuit.

[pebe]

OK. I’ll deal with the source resistance first. The attached file shows in Fig 1 how the input to the SSR is wired, with the 12V pulse being fed through R1.

Fig 2 shows the equivalent circuit, with the 560ohms input resistance of the SSR replaced by R2. Note that as the resistors are of equal value, the steady state after the cap has charged will be 6V at point ‘B’. Not all the current through R1 is charging the cap, because some of it is going through R2. So we need to simplify the circuit in order to know the current flowing into the cap and charging it.

If we look back at the pulse source from point ‘B’, there are two resistors connected there; R1 connects from the 12v pulse whose output resistance will be very low, so if we ignore the DC voltage for the moment, R1 is effectively connected to ground. Point ‘B’ is also connected to ground via R2, so R1 and R2 are effectively in parallel to provide the source resistance that charges C.

Thevenin’s theorem states, loosely, that in those circumstances “….the two resistors can be replaced by a single resistor equivalent to R1 and R2 in parallel, and feeding the cap from the open circuit voltage at point’B’…”

The open circuit voltage at point ‘B’ is 6V, so Fig 3 shows the equivalent circuit with point ‘B’ being fed from 6V through R3, which is 280 ohms. That is the circuit we can use for the CR time constant calculation.

To answer your last two questions:
If the input resistance of the SSR is not 560, the simplest way is to change R1 so it is the same as the actual SSR value. But 560ohms will be OK provided the steady state voltage at 'B' exceeds the SSR input threshold of 3V.
The cap can be any voltage over 6V

More to follow soon, about the time constant.....

[fizzy123]

That's so nice of you mate. I have downloaded the Yenka circuit simulator and can play around with different resistor and capacitor values. I read somewhere that using a transistor will eliminate the gradual increase in the voltage at the input and will only trigger when the desired voltage is achieved. Is there any benefit of using transistor for this application?

[pebe]

A transistor would decrease the switching time but it's not necessary because the SSR will itself be a fast switch. Also, you would need two transistors (a complimentary pair) because a single transistor would invert the pulse.

BTW, how are you getting on with Yenka, and do you still need the time constant explanation?

[fizzy123]

A transistor would decrease the switching time but it's not necessary because the SSR will itself be a fast switch. Also, you would need two transistors (a complimentary pair) because a single transistor would invert the pulse.

BTW, how are you getting on with Yenka, and do you still need the time constant explanation?

I understand the time constant a bit better now. Yenka is very good to test different values and combinations. I have a few resistors sitting at home, which have a Power rating of 250mW and I was worried about exceeding that. But with Yenka, I can exactly know that how I am going to achievce 560ohms with those without going over 50% of power rating.

I appreciate your help with this very much. Will update you on how I get on with putting together the circuit and testing.

[fizzy123]

Just to update that I put together the RC circuit and it has solved my problem. Pebe thanks for your help mate

[pebe]

I'm glad it all worked out OK.